With all the new token launches going to zero, principles to remember • A good project doesn't mean a good token • A project has 2 products: the product and the token • Project with revenue doesn't necessarily benefit token holders • Big names in the cap table + fomo-ed fundraising doesn't mean the token does well • Good product + narrative > tokenomics • Exchange listings are not catalysts anymore • Publicly published tokenomics information does not mean shit (if I have to explain to you...) • Low float + sky-high FDV ≠ real demand
What matters • Founders who pay attention to both the product and token • Teams who pay attention to a proper token launch (float, MMs, liquidity, CEX negotiations) • Founders that understand marketing, narrative, western + Asian GTM, KOL strategies • Founders that bother with IR and actively pitch to funds post TGE • Token flywheel, sinks and utility • Direct value-capture loop (fees, burns, buybacks) from the business • Effective community building • Low token inflation (token unlocks)
Robots are leaving the lab and coming on-chain. From @reborn_agi's “full-stack physical-AI enabler” to @PrismaXai “AWS for robot data".
I mapped 5 projects trying to own the machine economy, and possibly the trust gaps no one has solved yet.
- - - - -
1. Market topology Infrastructure rails: @PrismaXai (data-validation & tele-op hub) and @reborn_agi (Humanoid Data Platform) are racing to become the “AWS + Unreal Engine” of embodied AI (we love this term now over "robotics").
Community-first flywheels: whereas the likes of @frodobots (@BitRobotNetwork + @SamIsMoving) and @xmaquinaDAO are playing more towards the retail co-ownership narratives (memes, DAO treasuries) to bootstrap robotic fleets.
Open collaboration layer: @openmind_agi positions itself as an OSS runtime + P2P network that any robotics project can plug into, using blockchain rails and natural-language buses for alignment.
- - - - -
2. Competitive edges • Reborn AGI: hardware + sim moat • PrismaXAI: data is king; scale governance to oversee crowdsourced validators • Frodobots: viral growth, but hardware margins and safety liabilities (city regulations) loom • XMAQUINA: treasury diversification hedges risk, yet DAO execution (strategies/investments) will be decisive • OpenMind: wins mind-share with “shared intelligence” ethos, lowering integration friction via connectors
Projects still need to solve: • On-chain liability cover, safety scoring or compliance attestation • Inter-network liquidity and interoperability • Real revenue verticals
- - - - -
Seeing a lot of narrative plays and demos online and at @superai_conf but the robotics-crypto stack is definitely maturing to (crypto) economic primitives.
With OpenMind adding an open runtime & identity fabric, the next breakthrough will come from whoever secures trust, liquidity and interoperability across these robotic fleets.
A few @base ecosystem projects have been doing well lately
Think the reason for this IMO: • Compared to Solana, definitely fewer tokens = concentrated on things with fundamentals • Lesser attention + liquidity dilution • Projects will try to align with base = @coinbase listing hopium • People's edge in the Solana casino seems to be eroded, pivot to the Base casino with more PVE coins • @pumpdotfun token https://t.co/8SxFsVtNP9 which will cause a liq vacuum on Solana
Did a few minutes of research to keep up to date with the eco, some projects that are doing well: • @KeetaNetwork ($KTA)--> Visa-speed payment layer, wrapped in a KYC-ready L1 • @gizatechxyz ($GIZA) --> “Plug-in ML agents that trade, rebalance on-chain, use zkML for proves and trust-minimisation • @tigfoundation ($TIG) --> Turns hard optimisation puzzles into open bounties where the best algorithm wins (+ gets paid) • @Mamo_agent ($MAMO by @MoonwellDeFi) --> park USDC, grow yield, win prizes • @ReiNetwork0x ($REI) --> Bridges AI to on-chain (hard logic; verifiable intelligence) • @noicedotso ($NOICE) --> Attention is money, where every engagement (like, reply, or follow) becomes tokenised; social tipping
These are of course projs apart from the @virtuals_io ecosystem which I covered separately:
Narratives move faster than capital Enter InfoFi --> the primitive that lets you bet on both
TLDR InfoFi focuses on monetising attention, influence, and info, probably one of the more compelling narratives in crypto now.
Below I compare a few protocols: @KaitoAI, @cookiedotfun, @ethos_network, and @elfa_ai, each approaching the financialisation of digital mindshare from different angles.
These are platforms that structure and index high-signal info with an emphasis on measurable influence, data provenance, and actionable insights.
- - - - - -
1. Kaito ($KAITO) • The Bloomberg Terminal for crypto narratives, using AI to process thousands of CT, Reddit, and Telegram posts into searchable summaries ("Yaps") • Leaderboard system identifies top-performing accounts/wallets based on alpha calls • Creates a powerful social filter for crypto research
2. Cookie ($COOKIE) • Modular data layer for the agentic economy --> https://t.co/FVmBL8fOCG, tracks 1,500+ AI agents and their performance, powered by a backend called the Cookie DataSwarm. • This infra feeds into Agent Cookie, a super-agent providing social + on-chain intelligence
3. Ethos • Introduces a reputation graph for crypto, based on: Review, Vouch, Slash, and Influence. • Aim is to create a social trust layer that DAOs, protocols, and community mods can use to filter credible contributors or defend against Sybil attacks
4. Elfa • Sentiment engine indexing token mentions, KOL shills, and project/community updates in real time • Token mindshare is tracked hourly and Smart Accounts are analysed which is filtered by alpha potential
- - - - - -
Personal Take • Kaito dominates structured research workflows and integrations with projects • Cookie leads in infra for AI agents • Ethos builds foundational social trust infra for Web3 (reputation) • Elfa captures narrative volatility and CT pulse
Same with DEXes/L1s/L2s, all will have competition as protocols imitate and try to one up each other.
Useless - As absurd as fartcoin ("just a useless coin") - "I cant believe this useless coin went to $100m" is like akin fartcoin ripping - Orginial lore/meta - Letsbonk team incentivized to pump it + BonkGuy shilling it (main character arc)
@letsbonkdotai - AI/MCP exposure + working terminal with a prompt - Easy narrative: ai16z of Bonk eco - Intuitive chat interface + agent-managed wallet
Emerging meme and narrative tied to Believe (@believeapp) —> platform that enables internet-native fundraising and token launches.
TLDR • Decentralized fundraising and liquidity directly from internet-native companies/communities, without the need for VCs • Launching tokens instantly via social platforms (like how projects are now launching on Believe with one click) • A cultural shift where meme, momentum, and distribution > pitch decks and months of VC calls
Could spark the formation of a new asset class or market structure if traction picks up —> crypto-native, internet-native capital formation is now fast, public, and memetic.
Developments - Probably found a bottom at $2.6 after a huge retrace - Valuation hovering at the last round fundraise - No new supply hitting the market in the next 9 months - The large overhang is done (Boyco) - More projects to TGE + vibrant ecosystem (@InfraredFinance, @puffpaw_xyz, @EdenWeb3_Global, @KodiakFi) - As more activity takes place on-chain, bribes and fees increase --> compounds --> accrues back to main token - Smokey and team still actively on the ground talking to educate funds + strike partnerships with various teams - Always seem to get max attention, be it from a satire video to some contrarian project the team onboards
Think some things the team have to work: - Changing $BERA token narrative from just a pure farming token (pool 2) to getting it integrated more within the eco dApps - Still quite complicated for retails to understand as there are many yield sources - Potentially kingmake certain dApps and make sure they do well